In 2017, 3 UFO videos filmed by the US Navy were leaked to the media and then subsequently declassified by the Pentagon earlier this year. Already there's been quite a lot of analysis done on these and there are reasonable explanations for what these objects are if all we have to go by is the video footage. Along with the videos however, some of the pilots have given accounts of what was observed, most notably from David Fravor who has given several interviews with a compelling description of what was observed.
GOFAST is the title for one of these videos, which apparently shows an object moving fast above the surface of the water. On the History channel's 'Unidentified: Inside America's UFO Investigation', it was commented that the object is going at two thirds the speed of sound. It has been shown that whatever the object is, it is both not just above the water and not moving that fast - doing some maths on the numbers from the display readings will tell you this. Mick West has shown this quite clearly already but to over-engineer things, I've modelled the scene in Unity to simulate what happens in the video, the motion of both the aircraft and UFO over the duration filmed. This process also allows for better identifcation of the FOV (field of view of the camera) and thus the size of the object, as well as corroborating the parallax effect viewed.
The general steps involved in recreating the scene are simple:
place an Aircraft model or just an empty GameObject at the relevant height and in the centre of the scene (0x,7620y,0z) - doing this in meters and all metric units converted from the footage
place the camera at the same point as the aircraft and aim in the scene in the same direction as the video from the point it locks-on (L46,-23) and change the FOV to 1º.
place a waypoint 8148m in the look direction of the camera and also create a plane at that level, same scale as the water plane
move the aircraft to it's endpoint based on it's speed (calculated @369 knots from the readings), slight bank angle and the duration of the remainder of the video approx 21 seconds.
again aim the camera to the same angles as the video at the end of the scene. Fire a raycast from the camera towards the plane created in step 3 and create a second waypoint at the intersection. These 2 new Vectors are the start and end positions of the UFO. You can calculate speed from these vectors at this stage. remove the temporary plane.
linear interpolation over 21 seconds from the start to end points for: the camera angles, aircraft movement, UFO movement.
adjust FOV to match the motion of the ocean perceived in the video and adjust scale of object. Add overlay metrics as per video to check the stats broadly match up.
After doing the above, I have the object doing around 120knots @13,000 feet, which is not outside the realm of it being a small bird. I say small, because after reviewing the simulation at 1º FOV, the motion of the water is not fast enough to match what was in the video so either it was magnified further or NAR mode of the camera on the Raytheon ATFLIR is narrower than 1º; the more narrower the FOV, the smaller the object must be - in the case it might not even be 3 feet. I have seen documents of ATFLIRs that indicate you can magnify the image further once you're in the selected FOV mode, there are also documents that discuss NAR mode might be 0.7º on this particular model.
FOV comparison, 0.5º vs 1º:
There are some reasonable assumptions in the model e.g. constant altitude of both objects, somewhat linear movement etc and you can adjust all these easily but the end result will still show you the same bottom line. The object is actually moving the other way from its perceived motion by paralax.
The project folder can be downloaded from GitHub as well as just the standalone build, which has a couple of added features above what you can see in the video, such as removing the target lock to see how fast the UFO moves across the screen without it.
Metabunk discussion on GoFast:
Mick West's original video on analysing the GoFast video:
Commander David Fravor on the JRE podcast: